We will burn the planet expanding AI at the current technology, says NTT Data’s Khan

Scaling Artificial Intelligence(AI) at the speed at which consultants project is not possible by the laws of physics and may not be environmentally sustainable, said Tanvir Khan, who is the Executive Vice President and Chief Operating Officer of NTT DATA North America, part of the Japanese technology services and data centre company NTT Data, in an interview with The Hindu.

“AI is very compute intensive. So, it consumes electricity it consumes water. ..If you look at all the projections Gartner and others are making about AI, if we build all that computer over the next 20-30 years, we will burn the planet down. So, the projections of AI are not possible by the laws of physics,” Mr. Khan said while speaking about the environmental effects of the building data centres. Global business and technology company Gartner believes that artificial intelligence (AI) would touch a fourth of the information technology processes (IT). 

Mr. Khan feels that expanding AI will not be possible with the current technology and needs a fundamentally different know-how to scale. NTT Data, he said, had a photonic crystal-based technology called the Innovative Optical and Wireless Networks (IWON) which, he says, will help operating data centres water and energy efficient. 

Data centres are required to store, and compute large data required to run technology-intensive processes and become essential for the development of AI. This means that large amounts of cold water would be required to cool down the equipments that get heated up because of the high energy usage. Researchers and activists globally flag that this could suck up domestic water supply. India’s data centre story would reportedly double water consumption from the current 150 billion litre to 358 billion lite in 2030.  Currently, Mumbai is developing as a data centre hub and NTT Data Inc. will have its facility in Navi Mumbai in the coming 15 to 16 months, according to Mr.Khan.